Steve Reinhardt wrote:
> On Fri, Mar 26, 2010 at 10:32 PM, nathan binkert <[email protected]> wrote:
>   
>>> #1 sounds pretty good to me... 2 & 3 seem like overkill.  I don't
>>> think the overhead of the null pointer is that bad.
>>>       
>> Ok, cool.  I'll cook up a diff.
>>
>>     
>>>> I'm partial to doing 1, 3, or nothing (in that order.)  With #1, we
>>>> could also add a feature that requires the senderState to be popped at
>>>> every level and where we assert in the Packet destructor that
>>>> senderState is NULL.  This would ensure that you're not attaching
>>>> things to sender state and forgetting about them (causing either a
>>>> memory leak, or a programming error because you forgot to do
>>>> something.)
>>>>         
>> What about this feature where we ensure that senderState is NULL when
>> we call the destructor.  Think that's worth trying?  I could give that
>> a shot and see what happens.  Any reason not to do this?  (The
>> alternative is to try to delete the chain of sender state objects.)
>> Seems that one way or another we should try to avoid the memory leak.
>>     
>
> That seems reasonable to me.  You could always just put a warn() in if
> it's non-null and see if you hit it.
>
> Steve
> _______________________________________________
> m5-dev mailing list
> [email protected]
> http://m5sim.org/mailman/listinfo/m5-dev
>   
What about one way packets that don't collapse back to the sender? Do we
have any of those? Or do we always collapse back at least with the ack?

Gabe
_______________________________________________
m5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/m5-dev

Reply via email to